Estimation and Inference for Exponential Smooth Transition Nonlinear Volatility Models

31 Pages Posted: 18 Sep 2009

See all articles by Cathy W. S. Chen

Cathy W. S. Chen

Feng Chia University - Department of Statistics; Graduate Institute of Statistics & Actuarial Science, Feng Chia University

Richard H. Gerlach

University of Sydney

Boris Choy

University of Sydney Business School

Celine S. Y. Lin

Feng Chia University - Department of Statistics

Date Written: September 11, 2009

Abstract

A family of threshold nonlinear generalised autoregressive conditionally heteroscedastic models is considered, that allows smooth transitions between regimes, capturing size asymmetry via an exponential smooth transition function. A Bayesian approach is taken and an efficient adaptive sampling scheme is employed for inference, including a novel extension to a recently proposed prior for the smoothing parameter that solves a likelihood identification problem. A simulation study illustrates that the sampling scheme performs well, with the chosen prior kept close to uninformative, while successfully ensuring identification of model parameters and accurate inference for the smoothing parameter. An empirical study confirms the potential suitability of the model, highlighting the presence of both mean and volatility (size) asymmetry; while the model is favoured over modern, popular model competitors, including those with sign asymmetry, via the deviance information criterion.

Keywords: Asymmetric, Bayesian inference, Heteroskedastic, Markov chain Monte Carlo (MCMC), Normal scale mixtures distribution

JEL Classification: C11, C15, C22, C51, C52

Suggested Citation

Chen, Cathy W. S. and Gerlach, Richard H. and Choy, S. T. Boris and Lin, Celine S. Y., Estimation and Inference for Exponential Smooth Transition Nonlinear Volatility Models (September 11, 2009). Available at SSRN: https://ssrn.com/abstract=1471976 or http://dx.doi.org/10.2139/ssrn.1471976

Cathy W. S. Chen (Contact Author)

Feng Chia University - Department of Statistics ( email )

100 Wen Hwa Road
Taichung, 407
Taiwan
886 4 24517250 ext 4412 (Phone)
886 4 24517092 (Fax)

HOME PAGE: http://myweb.fcu.edu.tw/~chenws/

Graduate Institute of Statistics & Actuarial Science, Feng Chia University

100 Wenhwa Road
Talchung
Taiwan
886 4-24517250 ext 4412 (Phone)
886 4-2517092 (Fax)

HOME PAGE: http://myweb.fcu.edu.tw/~chenws/

Richard H. Gerlach

University of Sydney ( email )

Room 483, Building H04
University of Sydney
Sydney, NSW 2006
Australia
+ 612 9351 3944 (Phone)
+ 612 9351 6409 (Fax)

HOME PAGE: http://www.econ.usyd.edu.au/staff/richardg

S. T. Boris Choy

University of Sydney Business School ( email )

Cnr. of Codrington and Rose Streets
Sydney, NSW 2006
Australia

Celine S. Y. Lin

Feng Chia University - Department of Statistics ( email )

100 Wen Hwa Road
Taichung, 407
Taiwan

Register to save articles to
your library

Register

Paper statistics

Downloads
65
Abstract Views
470
rank
345,083
PlumX Metrics